For GLMs in the exponential family, we can obtain the standard errors for the regression coefficients as a function of the diagonal of the fisher information matrix. Does this still hold if the regression distribution is not in the exponential family (this is of course technically not a GLM but I'm not sure if there is a technical name for this kind of model)? For example, beta-binomial or dirchlet-multinomial? In this case, does it instead become necessary to use the diagonal of the Hessian?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.